feat: add file-backed canvas notes
Magic Notes previously supported rich-text records without a paged drawing or PDF annotation workflow. Add the integrated PeopleLib Fabric and Quill editor with flowing text, pen/highlighter, object transforms, images, paper templates, PDF backgrounds with native text extraction, and raster PDF export. Saves are manual; undo/redo remains scoped to an annotation page or Quill mode. Analyze drafts, saved canvases and canvas-source todos with page images and text when the default model supports images, or an explicitly labelled text-only fallback. Keep successful saves when automatic analysis fails and prevent MCP plain-text replacement of canvas records. Move entry bodies and binary assets into local note files with schema 38 migration and verified readback. SQLite retains metadata, indexes, revisions, todos and comments. Backups must coordinate SQLite and the adjacent notes directory while writes are paused or the app is closed; no one-click backup action is included. Show a continuous record stream with independent collapsible record index and AI pane, real first-page thumbnails, and scroll-only record navigation. Open a blank text/canvas composer by default and reset it after creation. Protect unsaved drafts and establish normalized editor baselines after saving so clean records do not prompt again when leaving. Validation: latest recorded focused run passed 149 tests across five files; typecheck, lint and git diff --check passed. The prior user-reported full run passed 4,713 tests before later corrections and UI changes; it is not a full validation of this final snapshot. Tests were not rerun for this commit, as requested; independent systematic review follows the commit. Release note: Create paged notes with handwriting, flowing text and PDF annotations, browse records through thumbnails, and request AI comments with an explicit text-only fallback. Existing notes migrate to local file storage; backups now need both the database and notes directory.
This commit is contained in:
mesalogo
@@ -1,11 +1,13 @@
|
||||
import { describe, expect, it } from 'vitest'
|
||||
import type { AgentExecutionRequest, AgentRuntime } from '../agent/runtime'
|
||||
import type { AgentExecutionRequest, AgentRuntime, RuntimeModelUsageEvent } from '../agent/runtime'
|
||||
import type {
|
||||
MagicNoteEntry,
|
||||
MagicNoteAnalysisOptions,
|
||||
MagicTodoItem
|
||||
} from '../../shared/magic-notes-contracts'
|
||||
import {
|
||||
analyzeMagicNoteEntry,
|
||||
analyzeMagicNoteDraft,
|
||||
analyzeMagicTodo
|
||||
} from './magic-note-analyzer'
|
||||
|
||||
@@ -21,6 +23,203 @@ const entry: MagicNoteEntry = {
|
||||
}
|
||||
|
||||
describe('magic note analyzer', () => {
|
||||
const png = 'iVBORw0KGgoAAAANSUhEUgAAAAEAAAABCAQAAAC1HAwCAAAAC0lEQVR42mP8/x8AAwMCAO+/l9sAAAAASUVORK5CYII='
|
||||
const canvasEntry = {
|
||||
...entry,
|
||||
content: {
|
||||
version: 2,
|
||||
kind: 'paged-canvas',
|
||||
pages: ['first', 'second'].map((id, index) => ({
|
||||
id, width: 800, height: 1000,
|
||||
background: { type: 'template', template: 'blank' }, objects: [{ text: index === 0 ? '目标' : '验收' }]
|
||||
})),
|
||||
assets: []
|
||||
},
|
||||
plainText: '第 1 页：目标\n第 2 页：验收'
|
||||
} satisfies MagicNoteEntry
|
||||
const options: MagicNoteAnalysisOptions = {
|
||||
requestId: '00000000-0000-4000-8000-000000000506',
|
||||
direction: 'general', format: 'structured',
|
||||
canvasImages: ['second', 'first'].map((pageId) => ({
|
||||
pageId, dataUrl: `data:image/png;base64,${png}`
|
||||
}))
|
||||
}
|
||||
|
||||
function recordingRuntime(error?: string) {
|
||||
const requests: AgentExecutionRequest[] = []
|
||||
const runtime: AgentRuntime = {
|
||||
requiresToolApproval: false,
|
||||
supportsToolExecution: false,
|
||||
async getStatus() {
|
||||
return { id: 'model', label: 'Test model', available: true, detail: 'Ready', supportsToolExecution: false }
|
||||
},
|
||||
async dispose() {},
|
||||
async *run(input: AgentExecutionRequest) {
|
||||
requests.push(input)
|
||||
if (error) {
|
||||
yield { requestId: input.requestId, type: 'error', status: 'failed', message: error } as const
|
||||
return
|
||||
}
|
||||
yield { requestId: input.requestId, type: 'text', delta: '{"comments":[{"kind":"summary","content":"明确目标与验收条件。"}]}' } as const
|
||||
yield {
|
||||
requestId: input.requestId, type: 'model-usage', callId: 'analysis-call',
|
||||
runtime: 'model', provider: 'test', model: 'test', inputTokens: 25,
|
||||
outputTokens: 10, cacheReadTokens: 0, cacheWriteTokens: 0
|
||||
} as const
|
||||
yield { requestId: input.requestId, type: 'done' } as const
|
||||
}
|
||||
}
|
||||
return { runtime, requests }
|
||||
}
|
||||
|
||||
const canvasTodo: MagicTodoItem = {
|
||||
id: '00000000-0000-4000-8000-000000000601',
|
||||
noteId: entry.noteId, entryId: entry.id, noteTitle: 'Canvas',
|
||||
sourceIndex: 0, source: 'note', title: 'Review release',
|
||||
instructions: 'Check acceptance criteria', completed: false,
|
||||
comments: [], revision: 0, createdAt: entry.createdAt, updatedAt: entry.updatedAt
|
||||
}
|
||||
|
||||
it.each([true, false])('analyzes canvas todos with task text and source context (vision=%s)', async (supportsImageInput) => {
|
||||
const { runtime, requests } = recordingRuntime()
|
||||
const comments = await analyzeMagicTodo(
|
||||
runtime, canvasTodo, options, undefined, undefined,
|
||||
{ supportsImageInput, content: canvasEntry.content }
|
||||
)
|
||||
expect(requests).toHaveLength(1)
|
||||
expect(requests[0]?.prompt).toContain(canvasTodo.title)
|
||||
expect(requests[0]?.prompt).toContain(canvasTodo.instructions)
|
||||
expect(requests[0]?.prompt).toContain('第 2 页：\\n验收')
|
||||
expect(requests[0]?.conversationId).toBe(`magic-todos:${canvasTodo.id}`)
|
||||
expect(requests[0]?.workMode).toBe('ask')
|
||||
if (supportsImageInput) {
|
||||
expect(requests[0]?.images).toHaveLength(2)
|
||||
} else {
|
||||
expect(requests[0]?.images).toBeUndefined()
|
||||
expect(requests[0]?.prompt).toContain('未查看页面图像')
|
||||
}
|
||||
expect(comments[0]?.inputMode).toBe(supportsImageInput ? 'canvas-images' : 'text-fallback')
|
||||
})
|
||||
|
||||
it('rejects missing canvas todo captures instead of pretending to see the source', async () => {
|
||||
const { runtime, requests } = recordingRuntime()
|
||||
await expect(analyzeMagicTodo(
|
||||
runtime, canvasTodo, { ...options, canvasImages: undefined }, undefined, undefined,
|
||||
{ supportsImageInput: true, content: canvasEntry.content }
|
||||
)).rejects.toThrow('截图缺失或与当前页面不匹配')
|
||||
expect(requests).toHaveLength(0)
|
||||
})
|
||||
|
||||
it('keeps legacy todo analysis limited to its title and instructions', async () => {
|
||||
const { runtime, requests } = recordingRuntime()
|
||||
const comments = await analyzeMagicTodo(
|
||||
runtime, canvasTodo, options, undefined, undefined,
|
||||
{ supportsImageInput: true, content: entry.content }
|
||||
)
|
||||
expect(requests[0]?.images).toBeUndefined()
|
||||
expect(requests[0]?.prompt).toContain(canvasTodo.instructions)
|
||||
expect(requests[0]?.prompt).not.toContain(entry.plainText)
|
||||
expect(comments[0]?.inputMode).toBe('text')
|
||||
})
|
||||
|
||||
it('passes complete composited pages in document order to AgentRuntime with page labels and text', async () => {
|
||||
const { runtime, requests } = recordingRuntime()
|
||||
const usage: RuntimeModelUsageEvent[] = []
|
||||
const comments = await analyzeMagicNoteEntry(
|
||||
runtime, canvasEntry, options, undefined,
|
||||
(event) => usage.push(event), { supportsImageInput: true }
|
||||
)
|
||||
expect(requests).toHaveLength(1)
|
||||
expect(requests[0]?.images).toEqual([
|
||||
{ name: 'page-1.png', mediaType: 'image/png', data: png },
|
||||
{ name: 'page-2.png', mediaType: 'image/png', data: png }
|
||||
])
|
||||
expect(requests[0]?.prompt).toContain('第 1 页 = page-1.png')
|
||||
expect(requests[0]?.prompt).toContain('第 2 页 = page-2.png')
|
||||
expect(requests[0]?.prompt).toContain('验收')
|
||||
expect(requests[0]?.workMode).toBe('ask')
|
||||
expect(comments[0]?.inputMode).toBe('canvas-images')
|
||||
expect(usage).toEqual([expect.objectContaining({ callId: 'analysis-call', inputTokens: 25 })])
|
||||
})
|
||||
|
||||
it('analyzes a visual-only draft using page images', async () => {
|
||||
const { runtime, requests } = recordingRuntime()
|
||||
const content = {
|
||||
...canvasEntry.content,
|
||||
pages: canvasEntry.content.pages.map((page) => ({ ...page, objects: [] }))
|
||||
}
|
||||
const comments = await analyzeMagicNoteDraft(runtime, '', options, undefined, undefined, { supportsImageInput: true, content })
|
||||
expect(requests[0]?.images).toHaveLength(2)
|
||||
expect(comments[0]?.inputMode).toBe('canvas-images')
|
||||
})
|
||||
|
||||
it('falls back to all extracted text for a text model, without silently slicing later pages', async () => {
|
||||
const { runtime, requests } = recordingRuntime()
|
||||
const content = {
|
||||
...canvasEntry.content,
|
||||
pages: canvasEntry.content.pages.map((page) => ({
|
||||
...page, objects: [{ text: '最后一页的结论' }]
|
||||
})),
|
||||
flow: { ops: [{ insert: '文'.repeat(30_001) }] }
|
||||
}
|
||||
const comments = await analyzeMagicNoteEntry(runtime, { ...canvasEntry, content }, options, undefined, undefined, { supportsImageInput: false })
|
||||
expect(requests[0]?.images).toBeUndefined()
|
||||
expect(requests[0]?.prompt).toContain('最后一页的结论')
|
||||
expect(requests[0]?.prompt).toContain('未查看页面图像')
|
||||
expect(comments[0]?.inputMode).toBe('text-fallback')
|
||||
})
|
||||
|
||||
it('rejects visual-only text fallback before issuing a request', async () => {
|
||||
const { runtime, requests } = recordingRuntime()
|
||||
const content = {
|
||||
...canvasEntry.content,
|
||||
pages: canvasEntry.content.pages.map((page) => ({ ...page, objects: [] })),
|
||||
flow: { ops: [{ insert: { image: 'placeholder' } }] }
|
||||
}
|
||||
await expect(analyzeMagicNoteEntry(runtime, { ...canvasEntry, content, plainText: ' ' }, options, undefined, undefined, { supportsImageInput: false })).rejects.toThrow('请切换支持图像输入的默认模型')
|
||||
expect(requests).toHaveLength(0)
|
||||
})
|
||||
|
||||
it.each([undefined, [], [options.canvasImages![0]!], [options.canvasImages![0]!, options.canvasImages![0]!]])('rejects missing or incomplete canvas captures (%j)', async (canvasImages) => {
|
||||
const { runtime, requests } = recordingRuntime()
|
||||
await expect(analyzeMagicNoteEntry(runtime, canvasEntry, { ...options, canvasImages }, undefined, undefined, { supportsImageInput: true })).rejects.toThrow('截图缺失或与当前页面不匹配')
|
||||
expect(requests).toHaveLength(0)
|
||||
})
|
||||
|
||||
it.each(['401 Unauthorized', 'network connection failed', 'context length exceeded'])('preserves provider errors without retrying text-only: %s', async (error) => {
|
||||
const { runtime, requests } = recordingRuntime(error)
|
||||
await expect(analyzeMagicNoteEntry(runtime, canvasEntry, options, undefined, undefined, { supportsImageInput: true })).rejects.toThrow(error)
|
||||
expect(requests).toHaveLength(1)
|
||||
expect(requests[0]?.images).toHaveLength(2)
|
||||
})
|
||||
|
||||
it('keeps ordinary notes text-only even when the model supports images', async () => {
|
||||
const { runtime, requests } = recordingRuntime()
|
||||
const comments = await analyzeMagicNoteEntry(runtime, entry, options, undefined, undefined, { supportsImageInput: true })
|
||||
expect(requests[0]?.images).toBeUndefined()
|
||||
expect(requests[0]?.prompt).toContain(entry.plainText)
|
||||
expect(comments[0]?.inputMode).toBe('text')
|
||||
})
|
||||
|
||||
it('includes native PDF text with its canvas page number during text fallback', async () => {
|
||||
const { runtime, requests } = recordingRuntime()
|
||||
const content = {
|
||||
...canvasEntry.content,
|
||||
pages: canvasEntry.content.pages.map((page, index) => ({
|
||||
...page,
|
||||
background: {
|
||||
type: 'pdf' as const, assetId: 'pdf-source',
|
||||
pageNumber: index + 10, text: `PDF 原生文字 ${index + 1}`
|
||||
}
|
||||
}))
|
||||
}
|
||||
await analyzeMagicNoteDraft(runtime, '', options, undefined, undefined, {
|
||||
supportsImageInput: false, content
|
||||
})
|
||||
expect(requests[0]?.prompt).toContain('第 2 页：\\nPDF 原生文字 2')
|
||||
expect(requests[0]?.images).toBeUndefined()
|
||||
})
|
||||
|
||||
it('uses ask mode and converts bounded JSON into comments only', async () => {
|
||||
let request: AgentExecutionRequest | undefined
|
||||
const runtime = {
|
||||
|
||||
Reference in New Issue
Block a user